MySQL Lists are EOL. Please join:

List:General Discussion« Previous MessageNext Message »
From:Mikael Fridh Date:November 6 2004 2:13pm
Subject:Re: replication log error
View as plain text  
Alvaro Avello wrote:
> 
>>mysql> show slave status\G;
>>*************************** 1. row ***************************
>>          Master_Host: hubble
>>          Master_User: replica
>>          Master_Port: 3306
>>        Connect_retry: 60
>>      Master_Log_File: hubble-bin.009
>>  Read_Master_Log_Pos: 646906124
>>       Relay_Log_File: gemini-relay-bin.008
>>        Relay_Log_Pos: 1006270724
>>Relay_Master_Log_File: hubble-bin.009
>>     Slave_IO_Running: No
>>    Slave_SQL_Running: No
>>      Replicate_do_db:
>>  Replicate_ignore_db:
>>           Last_errno: 0
>>           Last_error: error 'unexpected success or fatal error' on query 'INSERT
> INTO COBGE7╩łA'
>>         Skip_counter: 0
>>  Exec_master_log_pos: 646905758
>>      Relay_log_space: 1006271180
>>1 row in set (0.01 sec)
>>  
>> mysqlbinlog -f gemini-relay-bin.008 > nuevo.sql 
>>
>>and got :
>>
>>ERROR: Error in Log_event::read_log_event(): 'Event too small',
>>data_len=0,event_type=0
>>ERROR: Could not read entry at offset 909811060 : Error in log format or
>>read error
>>
>>

Try reading the master binlog, hubble-bin.009 and see if that file is 
also corrupt.
If it is not, issue a CHANGE MASTER TO on the slave using the position 
'Exec_master_log_pos' and 'hubble-bin.009'.

If the master log file is also corrupt, you could brute-forcibly find a 
good position in the binlog using something like the following...

i=646905758; while :; do mysqlbinlog --position $i hubble-bin.009; 
i=$(($i+1)); done

if it finds a good position it will start spitting log event entries, 
find the first good position and use that in your CHANGE MASTER TO...

saved me from a full disk binlog corruption on a Master server once.


-- 
  ___
|K  | Ongame E-Solutions AB - www.ongame.com
| /\| Mikael Fridh / Technical Operations
|_\/| tel: +46 18 606 538 / fax: +46 18 694 411
Thread
replication log errorAlvaro Avello5 Nov
RE: replication log errorAlvaro Avello5 Nov
  • Re: replication log errorMikael Fridh6 Nov